A central problem and budget issue for the federal government is how much to spend on public service advertisement campaigns (PSAs)? How much is viewership worth? This bears very much on the effectiveness of government outreach which will be more and more important for as a more open government explores ways to acheive participation and collaboration from citizens.
So first how unequal are the .Govs? Very.. Take a look at these visitor stats for the month of August..

Not all .Govs are create equal:

IRS.GOV	5,483,624
FUELECONOMY.GOV	3,054,628
DOL.GOV	1,209,052
CANCER.GOV	641,526
DRUGABUSE.GOV	81,646
HEALTHFINDER.GOV	60,425
NAUTICALCHARTS.GOV	311

But while some .Gov domains are woefully lacking in public awareness, there are others with plenty of awareness and regular visitors  as you can see from the table above.

So why cant the .Gov’s cross promote? Website visitors to .Gov domains are an under utilized natural digital resource. Just like the water on government lands which could be potentially used for hydropower or natural gas or coal.   Except there is no major environmental cost to its exploitation. (Unless of course you are a purist who will decry an anti-drug message soiling their NPS.gov landscape of old growth fonts – just a joke guys.. NPS.gov looks pretty hip!).

I propose a banner exchange system for .Govs websites (perhaps at first only federal .govs as that is easier to get started) which would allow cross promotion using a banner ad type system.   This would allow new public service campaigns, programs and other news to be promoted where there already is a lot of interest without having to start from scratch for every new program or campaign.

So here are some of the issues to be resolved. How many impressions does each agency get for its banner ads and how much mining of visitor information is allowed to determine which ads get delivered to whom.

A credit system would be better than purchase system to distribute ads. This way the agencies with bigger PR budgets and visitors rates would still have to give up space and couldn’t just purchase their own ads. In fact there should be minimum requirements for cross-agency promotion because, after all, that is the idea to expose audiences to new messages from agencies they don’t otherwise interact with.  Later on state and local .Govs might be able to become a part of the system but offering their own natural digital resources…visitors and webscape to the system.

Give agencies with larger number of visitors larger credits but give all agencies a minimum number of credits.   Like the senate and the house combined representation.  Texas gets more than Wyoming, but Wyoming still gets 3.    Notice I say agencies not domains.   If we give out credits based on the domain then it will start a gold rush of .gov domains which is not necessary going to improve the ability to get messages to the american people.

Of course the credit system doesn’t just depend on how many visitors you have but also on how valuable the landscape is that you are willing to give up. Most agencies probably would not give up a full banner ad but maybe a right side tower banner starting just above the “fold”.    (Sorry I am not up on all the official IAB terms). So the credits you earn would become adjusted based on relative value of the page real estate that you give up.  Again in a way which forces all agencies to give up significant real estate, so that the more popular sites can’t stick the less popular agencies in their footer.

Secondly, and here is the tricky part, do we tap into a commercial banner ad exchange system and get the rich demographic data which it can offer.  For those who don’t know, commercial banner ad systems can guess at demographics, income and other factors based on the behavior of a person as they go from site to site.  Of course in the government this opens up all sorts of privacy issues.  So in the short term, until these get resolved, use a simple IP lookup table to get an approximation of the Nielsen DMA or basically metro area for visitors. (All privately kept of course).   We can also use commercially available free statistics to determine the approximate demographic breakdown for each .gov domain so that agencies with targeted demographics can make informed decisions on where to place their banner ads.  Of course I am glossing over the complexity of different demographics for different sections or pages but lets crawl before we walk.

Bidding? Why not.
The exchange would basically allow agencies to bid with their credits on the metro areas and preferred .gov domain and pages they desired to make their ads seen at with the credits given and based on the outcome of the auction (think Google Adwords) get a proportion of the domains. I don’t think the bidding will get too crazy since a lot of federal messaging is geographically neutral and the demographics probably do not vary too much from site to site.   They would bid with their credits which they earn by giving up web real estate and having web visitors. Isn’t it great that getting ads in this systems is in itself and incentive to get more visitors on your website?

Reporting. Don’t forget centralized clickthrough reporting which hopefully has the ability to attach a conversion goal to it, such as a download, registration or video play. This will provide an additional set of visitor information to agencies deciding where to place ads and again be an incentive to give up good real estate as then people will bid higher on your spaces. Basically which visitors from which .Gov domains are more likely to make sure of your agency’s advertised information.

Is this moving forward? I know in the world of Web 2.0, this seems very Web 1.0. Then again some people still watch TV commercials too. Yes and its a bit oversimplified but I think its a start to a plan.  The proof will be in the clicks.

One more thing.. make sure these are no-follow links… don’t want the federal government to get in trouble with somebody with real power… Google!